PeterStuer

Surely I am missing something? right?

OpenAI ran a specific red team break out exercise in an environment that was not even air-gaped but connected to the open internet? It breached Hugging Face, and then Hugging Face is 'grateful for the collaboration'? wtf?

I see it more past 300-350k these days, but probably there is a smaller decline even past 250k.

All depends on context coherence. E.g. in Claude Code with the "1M context" models I am reluctant to push past 30%. So if that Codex 272K is 100% coherent than it is less of a difference. Still, even then the "hard fail" boundary has moved. Longer running agentic processes will surely run into this more often.

Anyone else silently added the "so you don't have to" staple GenAI title suffix?

Different era. At the time choosing the JVM and Java language for a multiplatform project that targetted a portable runtime was not controversial at all.

Regulatory capture. You get them to outlaw the part of the competion (safety!) that is unwilling to pricefix and participate in your margin and market division agreements.

"We could have designed our protocols to be minimally compatible with “a nation of laws,” but the tech bros insisted that compromise was treason, and, as a result, we will lose more privacy than necessary"

Unfortunately, no, you can't have a prophilactic that just makes you a little bit pregnant. We used to know this.

Somehow made me think of every 'modern' HR department.

Funny how these errors always go one direction.

The 'BOTH WAYS' in the original stood for 'both walking to the school and walking back from school'. ;)

Blaming this on Claude is a bit of a reach. The user very carefully goading Claude seems to be key here.

Also interesting buried a bit: user creats a website. Cloudflare immediatly erects a toll gate on it, without even asking.

I was also expecting a TUI based game.

Humanity has a proven infinite capacity for 'make work'. What that will mean for you depends on whether you will be lounging, poisoning and backstabbing in Capital City, or breaking large rocks into little rocks in the Districts for your next cup of porridge.

My first thought would be their server side extentions, code excecutoon sandboxes and document RAG search, being on by default? Probably should be an opt-in instead of an opt-out.

What is the minimum market share you need to make this work (a) assuming competition stays "neutral" and (b) assuming adversarial actions?

Next step? Maps Pro subscription 'blue' plan will guide you through the fastest routes, Maps Free users will be routed on the fast 'green' plan, but might ocasionally be diverted to other routes to 'balance' (free up blue routes') traffic under extreme congestion conditions.

Yep, the virtual universal fast toll road option.

"Ask the model" is usually the point where I realize that either what you need is not a specific answer to their specific question, but a tutor to guide you on a personal journey through the underlying fundamentals of more than one prerequisite areas of understanding, or, the question is too niche or domain specific that I would have to 'ask the model' as wel as I do not know. Now I did observe, not just with LLM but also with Google search, that many people struggle with this and never seem to get good results. They need meta-learning on how to ask the right questions, which I can offer, but often I notice at that point their eyes starting to glaze over as they want a 'direct' solution, me searching or llm guiding for them.

100% agree, with the slight nuance that the EU would not care 1-bit if the dad was scrolling EU propaganda.

I'm running PBS on a separate mini pc, it is the perfect forever near instant incremental backups that sold me. I did perfectly recover in 20 minutes after a full hardware failiure. Best recovery experience ever.

Let's hope it did not special case (overfit) for the specific tests. One of the failiure modes that in my experience takes the most effort to mitigate for.

"My main project right now is to establish a framework for large-scale, unsupervised code generation in our codebase... sifting through the unsupervised agent’s (Qwen’s) output"

Ouch. I love my local AI setup with Qwen, but that is a mismatch right there. That model is not the right match for that project. It's like trying to develop a major software solution by just throwing in hundreds of fresh junior programmers and have them spew out random code bits, while what you needed is a good PM, a great architect and a handfull of senior engineers. Might as well pack it in for a year until your model has grown into the ability for those rolls. There is a reason why Opus 4.6 and now Fable dramatically changed the SWE capabilities, and IMHO Qwen is not there yet.

Rather than 'plain' debian you can put it on Proxmox, and enjoy the bliss of Proxmox Backup Server.

I must admit lingering long since retired 'memories' are currently one of the biggest pitfalls of the setup. Wiping all 'memory.md' often leads to better sustain.

Have you read the DSA?

Sure we can nitpick semantics, but how would you describe the evolution in GenAI SWE capability over the last 18 months?

100% agreed, and I do experiment, but it is atm just a question of where to put the effort first.

My development platform of choice is Opus 4.8 --effort max. My 'in production' general model just switched from GPT-4o to GPT-5.2. I probably should test local inference for that part (some translation/knowledge extraction/summerization) next but it has not made it to the top of the priority list yet, even though 3 other small specialized task models in the app already run on prem.

Remember the Clipper Chip back in 1993 because personal computing was considered too dangerous a technologie to be accesible un-controlled/suprvised to the general public?

So a model named 'Fable 5' is "back". Both excited, as the previous Fable 5 I had access to for just 3 days was fantastic, and anxious as (yes, in my 1 person anecdote) the model refered to as 'Opus 4.8' was stealth nerfed over the last 2 weeks to a degree I had not experienced since the massive nerfs back in the GPT-3.x days. Fingers crossed.

Will they not just argue that you could share the assertion, and hence we need a 'trusted' verfication point to establish it is actually you in posession of the zkp token, right now. So turn on that smartphone camera right now and obediently follow our biometric verfication instructions ...